Der Vortrag von Prof. Dr. Olushina Olawale Awe (Alexander von Humboldt Scholar\ Guest Professor) findet statt am 26.11.2025, 14:15 - 15:45 Uhr, im Raum 5.211.
In today's data-driven society, artificial intelligence is playing an increasingly prominent role in shaping government, allocating resources, and engaging citizens. Yet as AI systems are deployed in civic contexts, from predicting voter turnout to targeting public health interventions, the need for transparency, interpretability, and public accountability becomes not just technical but deeply democratic. This talk presents a pedagogical approach to teaching interpretable AI models for classification and regression within the field of civic data science. Drawing on real-world datasets related to education, elections, urban infrastructure, and social equity, I explore how models used in my previous research, such as decision trees, support vector machines, rule-based classifiers, and interpretable regression techniques, can be used not only to extract patterns but also to support ethical reasoning and public understanding. I will also demonstrate how explanation tools like SHAP and LIME can be integrated into the learning process to reveal model logic and biases in accessible ways. The goal is to equip students, researchers, and civic professionals with the skills to build AI systems that are not only accurate but also explainable, inclusive, and socially responsive. This work underscores the value of aligning technical education with civic responsibility and preparing the next generation of data scientists to contribute to public life with integrity, clarity, and critical insight.
